locked
Heirarchy Design - Hot Site Requirement RRS feed

  • Question

  • I'm currently tasked with designing an SCCM heirarchy for our company. I've only been using SCCM 2012 for about a week in a lab environment to get used to it.  We currently have a single SCCM 2007 site at our headquarters. With 2012, I've been asked to setup a site at our headquarters (about 500 users), a hot site location, and 2 remote offices (20 users each). The hot site location has about 50 PCs on standby, with another 100 waiting to be imaged if needed. In the event that the headquarters building were to disappear, I need to be able to image PCs and push applications/updates at the hotsite and preferrably at the remote offices as well. What would be the best setup for this situation? I was told today by another co-worker that the only way to do this would be to set up a CAS, a primary site at headquarters, a primary site at the hot site, and a distribution point at each remote office. Is this correct or is there a more efficient way to manage this? Considering the requirements for a CAS, this seems like overkill to me, but to be honest, I don't know much yet. Would it be better to have 2 separate primary sites and just manually copy images/applications to the hotsite server as needed? Thanks for any input you can give.

    Thursday, March 7, 2013 2:28 AM

All replies

  • No. There are no true built-in hotsite capabilities in ConfigMgr and there are still a couple of single points of failure and using multiple primary sites and CAS has *nothing* to do with any type of resiliency; multiple primary sites are for scalability, communication challenges. and political requirements only. For what you've described, you'll have to think outside of the box and something like Hyper-v and it's site resiliency capabilities.

    Is there an SLA for this? What's the point of a hot failover for managing client systems? The backup and restore capabilities of ConfigMgr are very good and generally the main line of defense for most folks because not having ConfigMgr available for a few hours or even days has zero or very minimal business impact.


    Jason | http://blog.configmgrftw.com

    Thursday, March 7, 2013 4:40 AM
  • I'll recommend that you replace the hot-standby site idea with a stand-alone MDT solution. Software packages and Updates can be pushed out via a GPO (as a fallback plan). The idea of a hot-standby site is something I often hear as a request by never end up installing. Instead plan for multiple site system roles and make sure you test restore of ConfigMgr in production as part of your project.

    Kent Agerlund | My blogs: blog.coretech.dk/kea and SCUG.dk/ | Twitter: @Agerlund | Linkedin: Kent Agerlund | Mastering ConfigMgr 2012 The Fundamentals

    Thursday, March 7, 2013 9:37 AM
  • I am also currently involved in putting together a design and 'options' for a customer who also requires highly available and near disaster proof ConfigMgr infrastructure that is ready to deploy new systems, deploy software and patches etc.  We are planning to make MP and DP functionality as available as possible, but ultimately in certain circumstances a site recovery from backup has to occur onto replacement metal.

    I believe the evolution of ConfigMgr should be that it simply 'exists' in a fully distributed model without single points of failures, however it would then seem likely it would develop autonomy, build robots and take over the world.


    My Personal Blog: http://madluka.wordpress.com

    Thursday, March 7, 2013 11:47 AM
  • Is there an SLA for this? What's the point of a hot failover for managing client systems? The backup and restore capabilities of ConfigMgr are very good and generally the main line of defense for most folks because not having ConfigMgr available for a few hours or even days has zero or very minimal business impact.


    Jason | http://blog.configmgrftw.com

    Thanks Jason. The hot site location has about 50 PCs that are ready to go. If we have a situation where our HQ is going to be down for 48 hours or more, then we have to image and deploy the other 100 or so PCs to continue running the business. Our SLA is 4 hours to have a working solution ready. Complete building loss is obviously worst case scenario.

    So, just a primary at HQ and a DP at the hot sites and remote offices for imaging and updates?


    Thanks!

    Thursday, March 7, 2013 7:07 PM
  • I'll recommend that you replace the hot-standby site idea with a stand-alone MDT solution. Software packages and Updates can be pushed out via a GPO (as a fallback plan). The idea of a hot-standby site is something I often hear as a request by never end up installing. Instead plan for multiple site system roles and make sure you test restore of ConfigMgr in production as part of your project.

    Kent Agerlund


    Thanks Kent. I hadn't thought of a stand-alone MDT solution. I'll look into that as well.
    Thursday, March 7, 2013 7:08 PM